CN111915538B - Image enhancement method and system for digital blood vessel subtraction - Google Patents
Image enhancement method and system for digital blood vessel subtraction Download PDFInfo
- Publication number
- CN111915538B CN111915538B CN202010837429.1A CN202010837429A CN111915538B CN 111915538 B CN111915538 B CN 111915538B CN 202010837429 A CN202010837429 A CN 202010837429A CN 111915538 B CN111915538 B CN 111915538B
- Authority
- CN
- China
- Prior art keywords
- image
- subtraction
- noise reduction
- module
- enhancement
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 210000004204 blood vessel Anatomy 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000009467 reduction Effects 0.000 claims abstract description 68
- 238000012545 processing Methods 0.000 claims abstract description 16
- 230000002792 vascular Effects 0.000 claims abstract description 14
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 12
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 12
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 238000004458 analytical method Methods 0.000 claims abstract description 6
- 239000002872 contrast media Substances 0.000 claims description 22
- 238000007667 floating Methods 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000006835 compression Effects 0.000 claims description 7
- 238000007906 compression Methods 0.000 claims description 7
- 230000002194 synthesizing effect Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 230000002401 inhibitory effect Effects 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 5
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 238000002347 injection Methods 0.000 claims description 4
- 239000007924 injection Substances 0.000 claims description 4
- 230000000877 morphologic effect Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 238000011282 treatment Methods 0.000 abstract description 9
- 238000003384 imaging method Methods 0.000 abstract description 5
- 230000006378 damage Effects 0.000 abstract description 4
- 230000002829 reductive effect Effects 0.000 description 14
- 210000001519 tissue Anatomy 0.000 description 6
- 208000027418 Wounds and injury Diseases 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 239000003814 drug Substances 0.000 description 3
- 230000002708 enhancing effect Effects 0.000 description 3
- 208000014674 injury Diseases 0.000 description 2
- 238000001356 surgical procedure Methods 0.000 description 2
- 238000002560 therapeutic procedure Methods 0.000 description 2
- 206010002091 Anaesthesia Diseases 0.000 description 1
- 208000028399 Critical Illness Diseases 0.000 description 1
- 230000037005 anaesthesia Effects 0.000 description 1
- 238000002583 angiography Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002695 general anesthesia Methods 0.000 description 1
- 238000002697 interventional radiology Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 238000002690 local anesthesia Methods 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 210000004872 soft tissue Anatomy 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses an enhancement method and system for digital vascular subtraction, comprising a noise reduction module, a subtraction module, a detection module, an enhancement module and a synthesis module, wherein the digital vascular subtraction image is enhanced, the noise reduction module carries out noise reduction treatment on an original image and the subtraction image to improve the signal to noise ratio of the image, the subtraction module generates a subtraction mask and outputs a subtraction image, the noise reduction subtraction image is input into the detection module for analysis and detection, a vascular structure area is searched, the enhancement module carries out contrast enhancement and dynamic range adjustment on the noise reduction image, and finally the noise reduction subtraction image and the enhancement image are synthesized in the synthesis module according to the vascular structure area, and finally the enhanced vascular subtraction image is output; the invention can obviously improve the image quality of the digital blood vessel subtraction image by a digital image processing means, greatly reduce the dosage required by subtraction imaging and reduce the damage to a patient.
Description
Technical Field
The present invention relates to the field of medical devices, and more particularly to a method and system for digital angiogram enhancement.
Background
Interventional radiology, also known as interventional therapeutics, is an emerging discipline that has been rapidly developed in recent years, incorporating imaging diagnosis and clinical treatment. Under the guidance and monitoring of digital subtraction angiography machine, CT, ultrasonic and magnetic resonance imaging equipment, the special instrument is led into the lesion site of human body through natural duct or tiny wound of human body by using puncture needle, catheter and other interventional equipment to make minimally invasive treatment. Three clinical main-pillar disciplines have been developed in parallel with traditional medicine and surgery. The advantages of interventional therapy over surgical therapy are: (1) the utility model can complete the treatment without operation, without wound or skin incision of a few millimeters, and has small wound; (2) most patients only need local anesthesia rather than general anesthesia, so that the anesthesia risk is reduced; (3) the damage to normal tissues is small, the recovery is quick, and the hospitalization time is short; (4) intervention is also a good treatment for patients with advanced critically ill or without surgical opportunities who cannot tolerate surgery. Digital vascular subtraction (DSA) is an important function in interventional operations, can enable soft tissue blood vessels in a human body to be developed under X rays, is a necessary function in most interventional operations, but the X rays are harmful to the human body, and the traditional DSA process generally adopts X-ray irradiation imaging with larger dose, so that the development of a new technology has a certain practical significance in improving the image quality and reducing the X-ray dose.
Disclosure of Invention
The invention aims to: in order to overcome the defects of the prior art, the invention provides a method for enhancing a digital angio-subtraction image, which solves the problems of low image quality and high X-ray dosage in the prior art, and also provides a system for enhancing the digital angio-subtraction image.
The technical scheme is as follows: in one aspect, the present invention provides an image enhancement method for digital vessel subtraction, comprising:
(1) The method comprises the steps of collecting a contrast image under X rays as an original image sequence by controlling an X-ray perspective machine and a high-pressure injection pump to cooperatively work, carrying out time domain noise reduction on the original image sequence, and outputting a noise reduction image;
(2) Setting the noise reduction images of the previous frames as mask images, and carrying out logarithmic subtraction operation on the noise reduction images of the previous frames and the real-time images after noise reduction of the subsequent frames so as to obtain subtraction images;
(3) Carrying out frequency domain noise reduction on the subtraction image to obtain a noise reduction subtraction image;
(4) Carrying out normalization processing on the noise reduction image, stretching gray values of the biological tissue image in an interest gray range of a specific brightness range according to clinical experience setting parameters, inhibiting gray compression of a background or non-interest region gray range region, realizing contrast enhancement of the image, carrying out multi-resolution image decomposition on the image, carrying out different enhancement processing on different detail level information of the image under different resolutions, and synthesizing the image under different resolutions to obtain an enhanced image;
(5) Searching suspected blood vessel areas in the noise reduction subtraction image through gray threshold value processing and morphological analysis, marking the areas and recording suspected degrees;
(6) And according to the suspected blood vessel degree of each pixel on the image, combining the enhanced image and the subtraction image in a pixel-by-pixel proportion, and outputting the enhanced subtraction image or outputting the contrast image enhanced by the blood vessel region.
Further, the method comprises the steps of:
in the step (1), performing noise reduction in a time domain on the original image sequence, and outputting a noise-reduced image includes:
the original image sequence is circularly stored, the image exceeding the storage space covers the oldest image, each frame of image in the stored image sequence is multiplied by the weights and added, then the result is divided by the sum of all weights, the weight of each frame is decreased according to the time, the time domain noise reduction of the image is completed, and the noise reduction image is output.
Further, the method comprises the steps of:
the step (2) specifically comprises:
and adding 1 to the corresponding pixel gray values of the real-time image and the mask image, performing logarithmic operation, multiplying a coefficient obtained through a first test, wherein the prior coefficient is a floating point number manually selected after passing the test, performing logarithmic subtraction with the real-time image after subsequent noise reduction, namely alpha Log [ (A+1)/(B+1) ], and sequentially performing operation on each corresponding pixel gray value on the image to obtain a subtraction image.
Further, the method comprises the steps of:
the step (3) specifically comprises:
and manually selecting a local background image without contrast agent from the obtained subtraction image, carrying out Fourier transform, finding out a frequency component with noise being dominant in a frequency domain, recording the frequency domain position of the frequency component, and multiplying the frequency component by a floating point number between 0 and 1 on the frequency domain so as to realize attenuation of noise signals.
Further, the method comprises the steps of:
in the step (4), the gray value of the biological tissue image in the interest gray range in the specific brightness range is stretched, and the gray compression is performed in the gray range area of the background or non-interest area, so as to realize the contrast enhancement of the image, which specifically comprises:
constructing a specific S-shaped mapping curve, wherein the curve requires: the curve is smooth and continuous with more than 1 order in the definition range 0 to 1, the value range is between the parameters beta and 1, beta is a floating point number between 0 and 1, the slope of the curve is larger than 1 near 0, the slope is smaller than 1 near 1, the dynamic range is stretched and the lowest brightness is improved when the slope of the low gray value interval of the contrast agent with high probability is larger than 1, and the dynamic range is compressed when the slope of the high gray value interval of the contrast agent with high probability is smaller than 1.
Further, the method comprises the steps of:
the step (5) comprises:
constructing a hessian matrix to obtain characteristic values, namely obtaining characteristic values of the following matrix:
wherein I is xx ＝G xx *I,I xy ＝G xy *I,I yy ＝G yy *I,
g (x, y) is a two-dimensional Gaussian function, I is an image
The two eigenvalues are:
according to 2 eigenvalues lambda 1 ,λ 2 The suspected degree of the blood vessel under the scale is obtained according to the Franagi blood vessel similarity filter, and then the maximum value under each scale is taken as the final suspected degree of the blood vessel of the pixel, namely:
wherein p is a pixel point, sigma is a Gaussian scale factor, and beta, c and gamma are coefficients.
On the other hand, the invention also provides an image enhancement system for digital blood vessel subtraction, which comprises a noise reduction module, a subtraction module, a detection module, an enhancement module and a synthesis module, wherein the noise reduction module is responsible for reducing noise of an original image sequence to improve the signal to noise ratio of an image, the subtraction module is responsible for producing a subtraction mask and a subtraction image, the detection module is responsible for detecting a suspected blood vessel structure area in the subtraction image, the enhancement module is responsible for adjusting the dynamic range and the contrast enhancement of the image, and the synthesis module synthesizes the enhancement image and the subtraction image into a final enhancement blood vessel subtraction image according to the detected blood vessel structure.
The beneficial effects are that: compared with the prior art, the invention has the remarkable advantages that: the invention obtains high-quality images through processing the original images, and has a certain practical significance for interventional therapeutics.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
fig. 2 is a block diagram of the system of the present invention.
Detailed Description
As shown in fig. 1, the present invention provides an image enhancement method for digital vessel subtraction, in one aspect, the present invention provides an image enhancement method for digital vessel subtraction, including:
(1) The method comprises the steps of collecting a contrast image under X rays as an original image sequence by controlling an X-ray perspective machine and a high-pressure injection pump to cooperatively work, carrying out time domain noise reduction on the original image sequence, and outputting a noise reduction image;
in the step (1), performing noise reduction in a time domain on the original image sequence, and outputting a noise-reduced image includes:
the original image sequence is circularly stored, the image exceeding the storage space covers the oldest image, each frame of image in the stored image sequence is multiplied by the weights and added, then the result is divided by the sum of all weights, the weight of each frame is decreased according to the time, the time domain noise reduction of the image is completed, and the noise reduction image is output.
(2) Setting the noise reduction images of the previous frames as mask images, and carrying out logarithmic subtraction operation on the noise reduction images of the previous frames and the real-time images after noise reduction of the subsequent frames so as to obtain subtraction images;
and adding 1 to the corresponding pixel gray values of the real-time image and the mask image, performing logarithmic operation, multiplying a coefficient obtained through a first test, wherein the prior coefficient is a floating point number manually selected after passing the test, performing logarithmic subtraction with the real-time image after subsequent noise reduction, namely alpha Log [ (A+1)/(B+1) ], and sequentially performing operation on each corresponding pixel gray value on the image to obtain a subtraction image.
(3) Carrying out frequency domain noise reduction on the subtraction image to obtain a noise reduction subtraction image;
and manually selecting a local background image without contrast agent from the obtained subtraction image, carrying out Fourier transform, finding out a frequency component with noise being dominant in a frequency domain, recording the frequency domain position of the frequency component, and multiplying the frequency component by a floating point number between 0 and 1 on the frequency domain so as to realize attenuation of noise signals.
(4) Carrying out normalization processing on the noise reduction image, stretching gray values of the biological tissue image in an interest gray range of a specific brightness range according to clinical experience setting parameters, inhibiting gray compression of a background or non-interest region gray range region, realizing contrast enhancement of the image, carrying out multi-resolution image decomposition on the image, carrying out different enhancement processing on different detail level information of the image under different resolutions, and synthesizing the image under different resolutions to obtain an enhanced image;
the method comprises the steps of stretching gray values of a biological tissue image in an interest gray range in a specific brightness range, and inhibiting gray compression of a background or non-interest region gray range region to realize contrast enhancement of the image, and specifically comprises the following steps:
constructing a specific S-shaped mapping curve, wherein the curve requires: the curve is smooth and continuous with more than 1 order in the definition range 0 to 1, the value range is between the parameters beta and 1, beta is a floating point number between 0 and 1, the slope of the curve is larger than 1 near 0, the slope is smaller than 1 near 1, the dynamic range is stretched and the lowest brightness is improved when the slope of the low gray value interval of the contrast agent with high probability is larger than 1, and the dynamic range is compressed when the slope of the high gray value interval of the contrast agent with high probability is smaller than 1.
(5) Searching suspected blood vessel areas in the noise reduction subtraction image through gray threshold value processing and morphological analysis, marking the areas and recording suspected degrees;
constructing a hessian matrix to obtain characteristic values, namely obtaining characteristic values of the following matrix:
wherein I is xx ＝G xx *I,I xy ＝G xy *I,I yy ＝G yy *I,
g (x, y) is a two-dimensional Gaussian function, I is an image
The two eigenvalues are:
according to 2 eigenvalues lambda 1 ,λ 2 The suspected degree of the blood vessel under the scale is obtained according to the Franagi blood vessel similarity filter, and then the maximum value under each scale is taken as the final suspected degree of the blood vessel of the pixel, namely:
wherein p is a pixel point, sigma is a Gaussian scale factor, and beta, c and gamma are coefficients.
(6) And according to the suspected blood vessel degree of each pixel on the image, combining the enhanced image and the subtraction image in a pixel-by-pixel proportion, and outputting the enhanced subtraction image or outputting the contrast image enhanced by the blood vessel region.
Namely, for a certain pixel point p (i, j), the corresponding blood vessel similarity degree f (i, j), the enhanced image gray value e (i, j) and the subtracted image gray value s (i, j) are adopted.
Enhanced subtracted image s' (i, j) =s (i, j) ×f (i, j);
angiographic images with enhanced vascular regions e' (i, j) =e (i, j) ×1-f (i, j)) +s (i, j) ×f (i, j)
Because the whole process is subjected to image noise reduction and enhancement treatment, the requirement on the quality of an original image is reduced, and in turn, the dosage required by imaging is reduced, and the injury to a patient is reduced.
As shown in fig. 2, in another aspect, the present invention further provides an image enhancement system for digital vessel subtraction, where the system includes a noise reduction module, a subtraction module, a detection module, an enhancement module, and a synthesis module, where the noise reduction module is responsible for reducing noise of an original image sequence to improve an image signal-to-noise ratio, the subtraction module is responsible for producing a subtraction mask and a subtraction image, the detection module is responsible for detecting a suspected vessel structure region in the subtraction image, the enhancement module is responsible for adjusting a dynamic range and contrast enhancement of the image, and the synthesis module synthesizes the enhancement image and the subtraction image into a final enhanced vessel subtraction image according to the detected vessel structure.
Firstly, by controlling the X-ray perspective machine and the high-pressure injection pump to work cooperatively, a contrast image under X-rays is collected as an original image sequence, the first frames of the obtained image sequence are mask original images without contrast agent, the subsequent images are real-time original images containing contrast agent, the mask original images are used for manufacturing a subtraction mask in subsequent processing, and the subtraction mask is generated by operation processing with the subsequent real-time images.
And secondly, inputting the original image sequence into a noise reduction module for recursion noise reduction, namely circularly storing the original image sequence, automatically replacing the oldest image by the subsequent image exceeding the storage space, and according to the superposition synthesis of a plurality of images which are arranged to be the nearest image output each time, performing weighted average according to the distance of time to finish the time domain noise reduction of the image and outputting the noise reduction image. For the original image sequence I (n) N=0, 1,2,..k, noise-reduced image I ( ′ k) ＝(I (k) *1+I (k-1) *α 1 +...+I (0) *α k )/(1+α 1 +...+α k ) Alpha is a floating point number between 0 and 1 as a noise reduction coefficient.
The acquired original image sequence contains a large amount of noise signals due to the characteristics of X-ray images, and due to the clinical characteristics of contrast images, the correlation in time exists between each frame in the image sequence, the noise signals can be reduced by carrying out recursive noise reduction in a time domain, the spatial domain resolution of the images is improved, the signal to noise ratio is improved, and the time domain blurring, namely the tailing phenomenon, generated by the recursive noise reduction has no great influence on the contrast images.
And inputting the noise reduction image into a subtraction module, generating a subtraction mask according to a plurality of frames before setting as a real-time mask image, namely adding 1 (preventing from dividing by 0) to the corresponding pixel gray values of the real-time image and the mask image, carrying out logarithmic operation, multiplying the real-time image and the mask image by a coefficient obtained by priori, obtaining the subtraction image, wherein the priori coefficient is a floating point number manually selected after passing the test, carrying out logarithmic subtraction with the subsequent real-time noise reduction image, namely alpha Log [ (A+1)/(B+1) ] replaces A-B, namely carrying out operation on each corresponding pixel gray value on the image in sequence.
The noise reduction image sequence is input into a subtraction module, the subtraction module firstly generates a subtraction template, then the subsequent images are subjected to logarithmic subtraction operation to generate subtraction images, unchanged components on the images are removed, the difference part generated by the flow of the contrast agent is highlighted, and the response curves of the X-ray images are not linearly related but exponentially related, so that the logarithmic subtraction is adopted to perform the image subtraction operation to generate the subtraction images, and the effect obtained by the calculation method is better.
And thirdly, inputting the subtraction image into a noise reduction module for frequency domain noise reduction, wherein the subtraction image background still contains a large amount of X-ray quantum noise, and the frequency of the noise is obviously different from the frequency of the contrast agent part image, so that the noise of the background image part can be restrained by a method of carrying out frequency domain conversion to remove part of frequency components and then inversely converting the frequency components back into a space domain, and the noise reduction subtraction image is obtained. And manually selecting a local background image without contrast agent from the pre-obtained sample image, carrying out Fourier transform, finding out a frequency component with noise being dominant in a frequency domain, recording the frequency domain position of the frequency component, calling the information when noise is reduced, and multiplying the frequency component by a floating point number between 0 and 1 on the frequency domain, thereby realizing the attenuation of noise signals.
And thirdly, inputting the noise-reduced image into an enhancement module, carrying out gamma adjustment and multi-resolution enhancement on the noise-reduced image, adjusting the dynamic range and enhancing the contrast, carrying out normalization processing by analyzing the dynamic range of the image, namely carrying out histogram analysis on the image, defining the darkest pixel gray value as 0, defining the brightest pixel gray value as 1.0, and carrying out linear transformation on other pixel values. According to clinical experience setting parameters, stretching gray values of biological tissue images in a gray range of interest in a specific brightness range, and inhibiting gray compression of a gray range area of a background or a non-interest area, so as to realize contrast enhancement of the images, namely constructing a specific S-shaped mapping curve, wherein the curve requires: the method comprises the steps of stretching the dynamic range of a curve in a low gray value interval slope of a contrast agent with a large probability of being greater than 1 and improving the minimum brightness, compressing the dynamic range of the curve with a high gray value interval slope of the contrast agent with a large probability of being less than 1, decomposing images with multiple resolutions, carrying out different enhancement treatments on different detail level information of the images with different resolutions, and synthesizing the images with different resolutions to obtain enhanced images, wherein the slope of the curve is greater than 1 near 0 and the slope is less than 1 near 0, the dynamic range of the curve is stretched and the minimum brightness is improved in a low gray value interval slope of the contrast agent with a large probability of being greater than 1, the dynamic range of the curve is compressed in a high gray value interval slope of the contrast agent with a large probability of being less than 1, and then carrying out different enhancement treatments on the images with different resolutions to obtain the enhanced images.
Thirdly, inputting the noise reduction subtraction image into a detection module, wherein the contrast agent part on the subtraction image has certain regularity in image gray value and morphology, namely the blood vessel region marked by the contrast agent is similar to a certain tubular structure, the suspected blood vessel region can be searched through gray threshold processing and morphological analysis, and the regions are marked and the suspected degree is recorded, namely by settingA filter is calculated, the possibility of a tubular structure existing near each pixel is calculated, and the possibility is normalized to a numerical value between 0 and 1, and the numerical value is taken as the basis of the subsequent composite image, and the specific method is as follows: firstly, carrying out multi-scale decomposition on an image, and then solving three Gaussian second-order partial differential convolution values g on each scale on the image xx 、g xy And g yy Constructing a hessian matrix, solving the eigenvalues, solving the vascular suspected degree under the scale according to the Franagi vascular similarity filter according to the 2 eigenvalues, and then taking the maximum value under each scale as the final vascular suspected degree of the pixel.
Constructing a specific S-shaped mapping curve, wherein the curve requires: the curve is smooth and continuous with more than 1 order in the definition range 0 to 1, the value range is between the parameters beta and 1, beta is a floating point number between 0 and 1, the slope of the curve is larger than 1 near 0, the slope is smaller than 1 near 1, the dynamic range is stretched and the lowest brightness is improved when the slope of the low gray value interval of the contrast agent with high probability is larger than 1, and the dynamic range is compressed when the slope of the high gray value interval of the contrast agent with high probability is smaller than 1.
Further, the method comprises the steps of:
the step (5) comprises:
constructing a hessian matrix to obtain characteristic values, namely obtaining characteristic values of the following matrix:
wherein I is xx ＝G xx *I,I xy ＝G xy *I,I yy ＝G yy *I,
g (x, y) is a two-dimensional Gaussian function, I is an image
The two eigenvalues are:
according to 2 eigenvalues lambda 1 ,λ 2 The suspected degree of the blood vessel under the scale is obtained according to the Franagi blood vessel similarity filter, and then the maximum value under each scale is taken as the final suspected degree of the blood vessel of the pixel, namely:
wherein p is a pixel point, sigma is a Gaussian scale factor, and beta, c and gamma are coefficients.
And analyzing and detecting the noise reduction subtraction image, searching for a suspected vascular structure region, removing a non-vascular structure region, and avoiding the background enhancement of the subtraction image in the subsequent enhanced image synthesis process so as to improve the signal-to-noise ratio of the subtraction image.
And finally, inputting the noise reduction subtraction image, the blood vessel structure and the enhanced image into a synthesis module, and carrying out synthesis processing on the enhanced image of the subtraction image domain according to the suspected degree of the suspected blood vessel region, namely, synthesizing the enhanced image and the subtraction image in a pixel-by-pixel proportion according to the suspected blood vessel degree of each pixel on the image, and outputting the enhanced subtraction image or outputting the contrast image enhanced by the blood vessel region. Namely, for a certain pixel point p (i, j), the corresponding blood vessel similarity degree f (i, j), the enhanced image gray value e (i, j) and the subtracted image gray value s (i, j) are adopted.
Enhanced subtracted image s' (i, j) =s (i, j) ×f (i, j);
angiographic images with enhanced vascular regions e' (i, j) =e (i, j) ×1-f (i, j)) +s (i, j) ×f (i, j)
Because the whole process is subjected to image noise reduction and enhancement treatment, the requirement on the quality of an original image is reduced, and in turn, the dosage required by imaging is reduced, and the injury to a patient is reduced.
Claims (5)
1. An image enhancement method for digital vessel subtraction, comprising the steps of:
(1) The method comprises the steps of collecting a contrast image under X rays as an original image sequence by controlling an X-ray perspective machine and a high-pressure injection pump to cooperatively work, carrying out time domain noise reduction on the original image sequence, and outputting a noise reduction image;
(2) Setting the noise reduction images of the previous frames as mask images, and carrying out logarithmic subtraction operation on the noise reduction images of the previous frames and the real-time images after noise reduction of the subsequent frames so as to obtain subtraction images; the method specifically comprises the following steps:
the corresponding pixel gray values of the real-time image and the mask image are added by 1 and then subjected to logarithmic operation, and then multiplied by a coefficient obtained by priori, wherein the priori coefficient is a floating point number manually selected after passing a test, and the floating point number is subjected to logarithmic subtraction with the real-time image after subsequent noise reduction, namely alpha Log [ (A+1)/(B+1) ], and each corresponding pixel gray value on the image is subjected to operation in sequence to obtain a subtraction image;
(3) Carrying out frequency domain noise reduction on the subtraction image to obtain a noise reduction subtraction image;
(4) Carrying out normalization processing on the noise reduction image, stretching gray values of the biological tissue image in an interest gray range of a specific brightness range according to clinical experience setting parameters, inhibiting gray compression of a background or non-interest region gray range region, realizing contrast enhancement of the image, carrying out multi-resolution image decomposition on the image, carrying out different enhancement processing on different detail level information of the image under different resolutions, and synthesizing the image under different resolutions to obtain an enhanced image;
(5) Searching suspected blood vessel areas in the noise reduction subtraction image through gray threshold value processing and morphological analysis, marking the areas and recording suspected degrees; comprising the following steps:
constructing a hessian matrix to obtain characteristic values, namely obtaining characteristic values of the following matrix:
wherein I is xx ＝G xx *I,I xy ＝G xy *I,I yy ＝G yy *I,
g (x, y) is a two-dimensional Gaussian function, I is an image
The two eigenvalues are:
according to 2 eigenvalues lambda 1 ,λ 2 The suspected degree of the blood vessel under the scale is obtained according to the Franagi blood vessel similarity filter, and then the maximum value under each scale is taken as the final suspected degree of the blood vessel of the pixel, namely:
wherein p is a pixel point, sigma is a Gaussian scale factor, and beta, c and gamma are coefficients;
(6) And according to the suspected blood vessel degree of each pixel on the image, combining the enhanced image and the subtraction image in a pixel-by-pixel proportion, and outputting the enhanced subtraction image or outputting the contrast image enhanced by the blood vessel region.
2. The image enhancement method for digital vessel subtraction according to claim 1, wherein in the step (1), the time-domain denoising of the original image sequence, outputting a denoised image comprises:
the original image sequence is circularly stored, the image exceeding the storage space covers the oldest image, each frame of image in the stored image sequence is multiplied by the weights and added, then the result is divided by the sum of all weights, the weight of each frame is decreased according to the time, the time domain noise reduction of the image is completed, and the noise reduction image is output.
3. The image enhancement method for digital vessel subtraction according to claim 1, wherein said step (3) specifically comprises:
and manually selecting a local background image without contrast agent from the obtained subtraction image, carrying out Fourier transform, finding out a frequency component with noise being dominant in a frequency domain, recording the frequency domain position of the frequency component, and multiplying the frequency component by a floating point number between 0 and 1 on the frequency domain so as to realize attenuation of noise signals.
4. The image enhancement method for digital vascular subtraction according to claim 1, wherein in the step (4), the gray values of the biological tissue image in the gray range of interest in the set brightness range are stretched, and the gray compression of the background or non-region gray range region is suppressed, so as to enhance the contrast of the image, and specifically comprising:
constructing an S-shaped mapping curve, wherein the curve requires: the curve is smooth and continuous with more than 1 order in the definition range 0 to 1, the value range is between the parameters beta and 1, beta is a floating point number between 0 and 1, the slope of the curve is larger than 1 near 0, the slope is smaller than 1 near 1, the dynamic range is stretched and the lowest brightness is improved when the slope of the low gray value interval of the contrast agent with high probability is larger than 1, and the dynamic range is compressed when the slope of the high gray value interval of the contrast agent with high probability is smaller than 1.
5. A system based on the image enhancement method for digital vessel subtraction as claimed in any one of claims 1-4, characterized in that the system comprises a noise reduction module, a subtraction module, a detection module, an enhancement module, a synthesis module, the noise reduction module being responsible for reducing noise of an original image sequence to improve the signal-to-noise ratio of the image, the subtraction module being responsible for producing a subtraction mask and a subtraction image, the detection module being responsible for detecting a suspected vessel structure region in the subtraction image, the enhancement module being responsible for adjusting the dynamic range and contrast enhancement of the image, the synthesis module being responsible for synthesizing the enhancement image and the subtraction image into a final enhanced vessel subtraction image based on the detected vessel structure.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010837429.1A CN111915538B (en) | 2020-08-19 | 2020-08-19 | Image enhancement method and system for digital blood vessel subtraction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010837429.1A CN111915538B (en) | 2020-08-19 | 2020-08-19 | Image enhancement method and system for digital blood vessel subtraction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111915538A CN111915538A (en) | 2020-11-10 |
| CN111915538B true CN111915538B (en) | 2024-03-19 |
Family
ID=73279844
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010837429.1A Active CN111915538B (en) | 2020-08-19 | 2020-08-19 | Image enhancement method and system for digital blood vessel subtraction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111915538B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114037626B (en) * | 2021-10-28 | 2025-02-14 | 上海联影医疗科技股份有限公司 | Vascular imaging method, device, equipment and storage medium |
| CN114170143A (en) * | 2021-11-11 | 2022-03-11 | 复旦大学 | Method for aneurysm detection and rupture risk prediction in digital subtraction angiography |
| CN113902746A (en) * | 2021-12-13 | 2022-01-07 | 北京唯迈医疗设备有限公司 | Method and system for extracting blood vessel guide wire in medical image, electronic device and medium |
| CN114638757B (en) * | 2022-03-17 | 2024-08-02 | 苏州工业园区智在天下科技有限公司 | Noise reduction method, device, terminal and computer readable storage medium |
| CN114581443B (en) * | 2022-05-06 | 2022-08-26 | 中科慧远视觉技术(北京)有限公司 | Image processing method and device, computer equipment and readable storage medium |
| CN119295431B (en) * | 2024-12-09 | 2025-02-25 | 成都航空职业技术学院 | Bridge crack detection method |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4444196A (en) * | 1982-06-21 | 1984-04-24 | Diagnostic Technology, Inc. | Digital intravenous subtraction angiography |
| CN101051384A (en) * | 2007-05-14 | 2007-10-10 | 华中科技大学 | Self adaptive reinforced digital image reducing blood vessle image |
| CN102346911A (en) * | 2010-07-28 | 2012-02-08 | 北京集翔多维信息技术有限公司 | Method for segmenting blood vessel in digital subtraction angiography (DSA) image sequence |
| CN102512188A (en) * | 2005-06-06 | 2012-06-27 | 拜耳医药股份有限公司 | X-ray arrangement for the image representation of an examination object and use of the X-ray arrangement |
| CN102663709A (en) * | 2012-04-28 | 2012-09-12 | 大连海事大学 | Vessel enhancement method for X-ray angiogram |
| CN103679801A (en) * | 2013-12-01 | 2014-03-26 | 北京航空航天大学 | Angiocarpy three-dimensional reconstruction method based on multi-view X-ray film |
| CN103876764A (en) * | 2013-11-21 | 2014-06-25 | 沈阳东软医疗系统有限公司 | Vascular imaging method and device |
| CN104156931A (en) * | 2014-09-04 | 2014-11-19 | 成都金盘电子科大多媒体技术有限公司 | Novel digital subtraction angiography method |
| CN104463815A (en) * | 2014-11-24 | 2015-03-25 | 东软集团股份有限公司 | DSA image generating method and system |
| CN106803241A (en) * | 2017-01-20 | 2017-06-06 | 深圳市安健科技股份有限公司 | The processing method and processing device of angiographic image |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8111895B2 (en) * | 2006-12-06 | 2012-02-07 | Siemens Medical Solutions Usa, Inc. | Locally adaptive image enhancement for digital subtraction X-ray imaging |
| US8542900B2 (en) * | 2007-03-08 | 2013-09-24 | Sync-Rx Ltd. | Automatic reduction of interfering elements from an image stream of a moving organ |
| DE102016201702A1 (en) * | 2016-02-04 | 2017-08-10 | Siemens Healthcare Gmbh | Method for determining a navigation image to be displayed, X-ray device, computer program and data carrier |
-
2020
- 2020-08-19 CN CN202010837429.1A patent/CN111915538B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4444196A (en) * | 1982-06-21 | 1984-04-24 | Diagnostic Technology, Inc. | Digital intravenous subtraction angiography |
| CN102512188A (en) * | 2005-06-06 | 2012-06-27 | 拜耳医药股份有限公司 | X-ray arrangement for the image representation of an examination object and use of the X-ray arrangement |
| CN101051384A (en) * | 2007-05-14 | 2007-10-10 | 华中科技大学 | Self adaptive reinforced digital image reducing blood vessle image |
| CN102346911A (en) * | 2010-07-28 | 2012-02-08 | 北京集翔多维信息技术有限公司 | Method for segmenting blood vessel in digital subtraction angiography (DSA) image sequence |
| CN102663709A (en) * | 2012-04-28 | 2012-09-12 | 大连海事大学 | Vessel enhancement method for X-ray angiogram |
| CN103876764A (en) * | 2013-11-21 | 2014-06-25 | 沈阳东软医疗系统有限公司 | Vascular imaging method and device |
| CN103679801A (en) * | 2013-12-01 | 2014-03-26 | 北京航空航天大学 | Angiocarpy three-dimensional reconstruction method based on multi-view X-ray film |
| CN104156931A (en) * | 2014-09-04 | 2014-11-19 | 成都金盘电子科大多媒体技术有限公司 | Novel digital subtraction angiography method |
| CN104463815A (en) * | 2014-11-24 | 2015-03-25 | 东软集团股份有限公司 | DSA image generating method and system |
| CN106803241A (en) * | 2017-01-20 | 2017-06-06 | 深圳市安健科技股份有限公司 | The processing method and processing device of angiographic image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111915538A (en) | 2020-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111915538B (en) | Image enhancement method and system for digital blood vessel subtraction | |
| CN109166161B (en) | Low-dose CT image processing system based on noise artifact suppression convolutional neural network | |
| Zhao et al. | A new approach for medical image enhancement based on luminance-level modulation and gradient modulation | |
| CN110163809A (en) | Confrontation network DSA imaging method and device are generated based on U-net | |
| CN106530236B (en) | Medical image processing method and system | |
| CN113034641B (en) | Sparse angle CT reconstruction method based on wavelet multi-scale convolution feature coding | |
| CN107845079A (en) | 3D shearlet medicine CT video denoising methods based on compact schemes | |
| CN104574284A (en) | Digital X-ray image contrast enhancement processing method | |
| CN115984394A (en) | A Low-Dose CT Reconstruction Method Combined with Prior Image and Convolutional Sparse Network | |
| Mredhula et al. | An extensive review of significant researches on medical image denoising techniques | |
| CN112258438B (en) | LDCT image recovery method based on unpaired data | |
| CN102663709B (en) | Vessel enhancement method for X-ray angiogram | |
| Thakur et al. | Medical image denoising using convolutional neural networks | |
| Bansal et al. | A proposed approach for biomedical image denoising using PCA_NLM | |
| CN114864095B (en) | Analysis method for narrow coronary artery blood circulation change under combination of multiple exercise intensities | |
| Fonseca et al. | X-ray image enhancement: A technique combination approach | |
| Zhang et al. | Medical image fusion based a densely connected convolutional networks | |
| CN109584322A (en) | Based on the smooth Shearlet medicine PET image denoising method of frequency domain direction | |
| Kalamkar et al. | Multi-modal medical image fusion using transfer learning approach | |
| CN110570369B (en) | Thyroid nodule ultrasonic image denoising method | |
| CN115731158A (en) | A Low Dose CT Reconstruction Method Based on Residual Domain Iterative Optimization Network | |
| CN110327066B (en) | Cardiac motion signal acquisition method, cardiac motion signal acquisition apparatus, computer device, and storage medium | |
| Ishfaq | A review on comparative study of image-denoising in medical imaging | |
| Ouahabi | Image Denoising using Wavelets: Application in Medical Imaging | |
| Hu et al. | A two-stage deep-learning framework for CT denoising based on a clinically structure-unaligned paired data set |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |